32
  • Why does The Definitive C++ Book Guide and List get a pass, but I can't ask my own book recommendation question?
  • Can I ask my own book recommendation question in a way that is acceptable to the community, like the C++ book list question?
13
  • 7
    asked nearly 4 years ago.. Aug 21, 2012 at 16:40
  • 1
    What if you just edit the existing C++ post for a new topic completely? That way you already have some answers to go through
    – random
    Aug 21, 2012 at 16:40
  • 1
    @MartijnPieters: It doesn't have a historical lock. We tried to lock it once, but the C++ community asked us to remove it so they could maintain it.
    – user102937
    Aug 21, 2012 at 16:41
  • 2
    @RobertHarvey: Perhaps someone has to put their foot down and lock it anyway? :-P Aug 21, 2012 at 16:42
  • @RobertHarvey Is this going to be a self-answered question or a real question?
    – Bart
    Aug 21, 2012 at 16:43
  • 11
    @MartijnPieters Locking it would be counter productive, since it could no longer be maintained. It either has to stay, or go.
    – Tim Post
    Aug 21, 2012 at 16:43
  • @Bart: I have my own thoughts, but I figured I'd let the community weigh in first.
    – user102937
    Aug 21, 2012 at 16:43
  • 3
    @TimPost: That's fine by me. It may be a great post, but it should be moved to a real Wiki and off StackOverflow; it's no longer on-topic here. Aug 21, 2012 at 16:44
  • 1
    This may quickly become faq-proposed since we appear to have added an exception for that particular question. @MartijnPieters That's exactly what Robert seems to be addressing here.
    – Tim Post
    Aug 21, 2012 at 16:45
  • 2
  • 6
    @MooingDuck: Didn't know it had been asked before. I was kinda hoping to get better feedback than what is on the older question. The C++ book list is aggressively maintained and curated by the C++ community, which is what makes it a good post. If there were some assurances that such curation would take place on other book lists, they might be more palatable. Alas, most questions like this turn into forum posts, with endless duplication and "me too" answers.
    – user102937
    Aug 21, 2012 at 16:48
  • 3
    Because other questions don't have enough recalcitrants to offset it being in the wrong place
    – random
    Aug 21, 2012 at 16:48
  • Related: meta.stackexchange.com/q/124635/167443
    – yoozer8
    Aug 21, 2012 at 17:22

3 Answers 3

23

The big problem with book recommendation posts - and similar questions - is that they tend to devolve into polls:

For me, and I think I speak for many others, the only time you really know you’re looking at a poll is when you see those one-line answers. But we can only close questions. Is it always the fault of the questioner that the answers suck? Do they deserve all the blame for the fact that people with nothing useful to say want to participate anyway and bring their reddit-style “tweets” into the answers?

That's a quote from a quote from Real Questions Have Answers, a blog post that heralded a revised FAQ and a much tighter focus on questions that can be constructively, usefully answered. The relevant message there, ensconced in the FAQ on every site, is this:

... avoid asking subjective questions where …

  • every answer is equally valid: “What’s your favorite ______?”
  • ...
  • there is no actual problem to be solved: “I’m curious if other people feel like I do.”
  • ...

Note that it is possible to have a question that asks for a list of things that does solve a real problem and doesn't treat every answer as valid. But it is very, very hard, and most questions of that sort degrade badly over time.

Why does this question get a pass, but I can't ask my own book recommendation question?

That question gets a pass because it solves a real problem identified by many - perhaps most - people in Stack Overflow's C++ community.

Can I ask my own book recommendation question in a way that is acceptable to the community, like the C++ book list question?

Probably not. If you have to ask, almost certainly not. That said, here are a few guidelines if you wish to attempt it anyway:

  • Identify your problem: "I'm looking for a good book to read on X" probably won't fly. Have you seen Amazon? They've built an amazing system for categorizing and rating books on just about every topic you can imagine. If you're looking for a book recommendation here, you'd better start with a clear description of why Amazon failed you.

  • Be specific: describe the type of information you're looking to garner. The C++ question isn't asking for every C++ book ever published. They're looking for the cream of the crop, and most importantly, they're looking for answers that explain why.

  • Be part of a healthy community willing to curate and update the results: Look at the revision history for that C++ question. See how they keep it updated? That's important, and you're not going to do it by yourself.

  • Be willing to fight for it: Again, look at that revision history. Chances are, folks are going to try to shut your question down. Be prepared to defend it, and calmly, clearly rebut charges that it is not constructive. This will be a lot easier if you're also part of that healthy community dedicated to keeping it up-to-date.

Sound hard? Like too much work to bother with? You're probably right, and that's the idea - when these things aren't hard-won labors of love, they quickly become dung heaps of obsolete and spammy no-content responses. If it's worth doing at all, it's worth doing well.

25
  • 1
    The boilerplate needs to be a boilerchassis for this post notice distinct
    – random
    Aug 21, 2012 at 17:00
  • I'm upvoting only for the word "ensconsed". And maybe because I agree with it too. Aug 21, 2012 at 17:08
  • 1
    Sure, but are you guys going to implement something for all tag wikis like Jerry described? I think that could be a great addition to the site. Aug 21, 2012 at 17:08
  • @Adam: right now, I don't know. We've talked about this - endlessly - and the problem is that to make tag wikis useful for this sort of thing, we're putting a lot of effort into 1) recreating a lot of functionality that already exists on normal questions (for all the crap associated with it, CW actually works pretty well for actual wikis), and would only be used for a tiny handful of posts.
    – Shog9
    Aug 21, 2012 at 17:10
  • 2
    for all the crap associated with it, CW actually works pretty well for actual wikis - agree. So what can we do to allow book-style CW questions without hoards and hoards of close and delete votes rushing in? Aug 21, 2012 at 17:11
  • @Adam: that, would make a good question all on its own...
    – Shog9
    Aug 21, 2012 at 17:16
  • 1
    @Shog9 - YOU ask that question - I'm not dealing with the crap that'd come in — looking at you, random... Aug 21, 2012 at 17:17
  • 1
    @AdamRackis I think the ideal solution is to make a place for them. The problem is getting it visibility. This whole business of curating such a list so the 'best' recommendations find their way into the question is simply not tenable. The CW book list is a bad idea. If we're going to allow it, we need to properly support it.
    – Tim Post
    Aug 21, 2012 at 17:17
  • @Tim - maybe. Tag wikis can be left to whither and die, too. I think being more liberal in what gets allowed with CW wouldn't be too terrible. If posts are mediocre and turn into polls, then it won't get many votes and won't show up on top questions and such. Aug 21, 2012 at 17:19
  • 4
    @Tim: keep in mind, even if we did carve out a separate area for these... There's negligible value in re-implementing Amazon reviews, badly.
    – Shog9
    Aug 21, 2012 at 17:19
  • 6
    There are actually a few other questions similar to that one, @Adam. They're incredibly rare gems though - you can dig up your backyard all day long and never find a Hope Diamond in it. Truth is, a few wikis are useful - but a lot of them are just folks thinking, "wouldn't it be great if I had a list of X - maybe I can get others to build it for me".
    – Shog9
    Aug 21, 2012 at 17:23
  • 7
    The question in question fails on all of your points: 1) Amazon has reviews and indexes C++ books as well 2) at 3 pages of 50 answers per page, the answers are not pruned very effectively. There's a ton of crap there and most of it is one-liners. The question is groomed well though, but that's not a saving grace 3) They need to be more aggressive about maintaining the answers since those are just noise 4) they'll defend it, but, and more often than not we've backed down because of idle threats, not because it was the right thing to do for the site.
    – casperOne
    Aug 21, 2012 at 17:37
  • 1
    Dealing with what kind of crap? You mean book lists without the Amazon UI? @ada
    – random
    Aug 21, 2012 at 18:17
  • 4
    @AdamRackis If posts are mediocre and turn into polls, then it won't get many votes and won't show up on top questions and such. That's not how polls work. The problem with polls was that they get crazy votes because a vote suddenly means "yeah, that's kinda cool" not "this is a complete, self contained useful answer". Granted they no longer get those votes because they get closed, downvoted & deleted due to general community consensus.
    – Ben Brocka
    Aug 21, 2012 at 18:21
  • 6
    @casperOne has hinted at what I suspect is at least part of the true answer: C++ users are a cantankerous bunch who tend to lead a separate existence except for occasionally descending on meta en masse and blasting us away for our ignorance and incomprehension :)
    – Benjol
    Aug 23, 2012 at 6:49
42

Given the (high) degree to which almost all programming depends upon books (online resources are rarely a complete substitute), what should probably be done is every tag wiki should have a tab (or some other convenient way to access a section) devoted to the book list for that subject.

I should probably add, however, that it would probably be perfectly fine/sensible for the tab/section to be named something relatively generic like "resources" instead of being specifically a "book list". There may be some languages, libraries, tools, etc., for which online resources really are good/useful/preferred, and I certainly have no intent to discriminate against such cases.

17
  • I honestly thought this was a feature-request already ...
    – jcolebrand
    Aug 21, 2012 at 16:50
  • 20
    Great idea :) Now how do we make that more visible? Tag wikis are like the member of the opposite sex that you only call when intoxicated, depressed or really bored.
    – Tim Post
    Aug 21, 2012 at 16:54
  • 2
    @TimPost: That is a problem, for sure. One thing that might help would be to add the tag wiki/book list to the list when you're voting to close a question as a duplicate. That (probably) won't prevent new users from asking questions that are already covered, but will at least make it easy to direct them there afterwards. Right now, it's much easier to direct them to another question than to the tag wiki. Aug 21, 2012 at 17:00
  • @Tim - I think that would solve itself if this feature were to be added. Aug 21, 2012 at 17:06
  • 1
    Re. the edit - or just let users arbitrarily add new tabs—books, frameworks, IDEs, etc. Aug 21, 2012 at 17:14
  • 1
    @AdamRackis: I considered that possibility, but suspect it's a fair amount of extra work. I'd rather see something usable now than something perfect in 10 years. Aug 21, 2012 at 17:15
  • 3
    I think this idea is the most sensible thing to explore. We can't just keep making 'special' exceptions with no clear objective method for arriving at them. Otherwise, it's 'why has my book list not fought hard enough?' city.
    – Tim Post
    Aug 21, 2012 at 17:26
  • @Jerry - Nick Craver could add this feature in about 20 minutes, I'm guessing... Aug 21, 2012 at 19:27
  • @AdamRackis we're waiting... Aug 21, 2012 at 19:33
  • I think this was tried for a couple of tags but people complained that the information was now "hidden"
    – ChrisF Mod
    Aug 21, 2012 at 19:37
  • @ChrisF: This question was merged into the tag wiki for a while, and no that didn't work (pretty much at all). It wasn't (sufficiently) accessible, which is crucial. I'm not sure adding a tab will be enough either, but I think the combination of a tab and easy linking to it from duplicates is likely to help at least some. Aug 21, 2012 at 19:45
  • It was hidden because no one wanted to link to it and because the URL slug doesn't give you a hint what you're pointing them to
    – random
    Aug 21, 2012 at 20:11
  • 2
    My problem with making a formal feature for this is it really dilutes what SE is about; answering practical, specific questions. It's already a Q&A site with a meta Q&A site plus a chat site, now we're talking about taping a book list/IDE list/link depository on there too. It's feature creep of the worst kind. Tags are for categorizing questions and answers, directly related to Q&A. Making them something else is just hammering a square peg in a round hole. Even if it works it's not really supposed to.
    – Ben Brocka
    Aug 23, 2012 at 12:37
  • 1
    @JerryCoffin those aren't questions, per our format. We can't consider those questions. Why? Because if we try to answer them, as you imply, we have to make something that's not an answer. Rather than trying to answer the unanswerable we're really supposed to close those questions.
    – Ben Brocka
    Aug 23, 2012 at 13:40
  • 1
    It's both, but they're unrelated issues. The problem isn't that the asker has no experience, but that the answers have no limits.
    – Ben Brocka
    Aug 23, 2012 at 14:53
1

There is a great need to have a list of most useful resources here.

However, we have tag wikis and we have StackOverflow blog, so such resouce lists should belong there.

I vote for locking that book recommendation question and moving the resouce list to the tag wiki or to the special blog entry.

1
  • Tag Wikis are not quite ready for that. Some ideas have been floated to improve them, but nothing definitive. See here for some additional background: meta.stackexchange.com/questions/122197/…
    – user102937
    Feb 13, 2014 at 18:18